人们在处理高光谱图像时一般要对一些典型地物进行光谱分析、特征波段的提取,以便提取出最大量的有效信息,剔除无用或冗余的信息,然后再进行分类识别.采用小波变换的分析方法,选用合适的小波进行分解,根据分解后的高频分量中包含的重要信息,利用局部相邻的正负极值点找出对应于原始光谱曲线上每个吸收带的左右边界;利用局部过零点,即可比较精确的提取出各个吸收带的中心波长.该方法比传统的光谱特征提取方法更简洁、有效,实验证明为一种比较理想的光谱特征提取方法.
参考文献
| [1] | Hsu Paihui,Tseng Yihsing.Multiscale analysis of hyper-spectral data using wavelet for spectral feature extraction [OL].Http://www.gisdevelopment.net. |
| [2] | Hu Changhua,Li Guohua,Liu Tao,et al.Wavelet Analysis-System Analysis and Design Based on MATLAB 6.X (the Second Edition) [M].Xidian University Publishing House,2004. |
| [3] | Bruce L M,Cliff K H,Jiang Li.Dimensionality reduction of hyper-spectral data using discrete wavelet transform feature extraction [J].IEEE Transactions on Geoscience and Remote Sensing,2002,40(10). |
| [4] | Mallat S.A theory for multi-resolution signal decomposition:the wavelet representation [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1989,11:674-693. |
| [5] | Http://www.bearcave.com/misl/misl_tech/wavelets/matrix/,A Linear Algebra View of the Wavelet Transform. |
| [6] | Http://www.bearcave.com/misl/misl_tech/wavelets/daubechies/,The Daubechies D4 Wavelet Transform. |
| [7] | Zhang Jingyuan,Zhang Bing,et al.Analysis of feature extraction methods based on wavelet transform [J].Signal Processing (信息处理),2000,16(2):156-162 (in Chinese). |
上一张
下一张
上一张
下一张
计量
- 下载量()
- 访问量()
文章评分
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%